 |
 |
 |
 |
| |
|||||
|
|
|
Home > News & Events > MITRE Publications > The MITRE Digest > | |||||||||||||||||||
Making Tomorrow Happen Today: DirAc and the New GPS November 2004
When warfighters must know their position and the precise time, their GPS receivers will soon deliver the goods using the new military signal—the M-code signal. But will receivers be able to find and lock in the M-code signal directly? MITRE's DirAc integrated circuit prototype answers that question today. Global radio navigation satellite system technology—and, in particular, the U.S.'s NAVSTAR Global Positioning System (GPS)—has become so commonplace (from car dashboards to locator tags in purebred dogs), that it's easy to forget the military origins of this valuable utility. The explosion in demand for civilian GPS capacity has been mirrored by growing military use of the system. These trends, along with the increasingly sophisticated jamming capabilities of our adversaries, have led the Department of Defense (DOD) to modernize GPS. In the late 1990s, the DOD decided to develop a new GPS military signal, the M-code, to enable improved civil GPS services while providing quantum increases in military capability. Many organizations, including MITRE, supported the GPS Joint Program Office in designing the new signal, which will first be transmitted in 2005. One of the challenges: Could receivers lock directly onto (or "directly acquire") the M-code signal, even when their internal clocks were off by fractions of a second or even seconds? The entire design team agreed that direct acquisition, which avoids use of so-called acquisition aids, was the best way to provide the capabilities stated in the requirements—but only if the basic technology could be ready in time. A quick overview of how GPS works illustrates why this challenge is important to overcome. GPS has three major components: the user equipment or receiver (held by a soldier or hiker or installed in systems and vehicles), the control segment (which monitors the signals from satellites and sends commands to the satellites), and the space segment (the 27 satellites themselves, each of which orbits the earth twice daily). Each satellite transmits signals for worldwide civilian use, as well as military signals. The current military signal, called P(Y) code, is encrypted to restrict its use to U.S. and allied forces. The civilian C/A-code is potentially jammable, but it provides the initial synchronization that the P(Y) code needs to acquire the right satellite signal. When receivers can directly acquire the same signal used for navigation, the control and space segments are simpler, and all available signal power (always a scarce commodity) can be used for navigation by the receiver. "A GPS receiver processes the satellite signals to determine your position," explains MITRE's John Fite, a principal software systems engineer. "A critical moment is signal acquisition, when the receiver first locks onto the signal from the satellite. It's an essential step, but also difficult to accomplish. Many existing military receivers acquire the signal using the civilian signals, which is easy to do but not preferred in electronic warfare environments. Directly acquiring the current military signal requires computationally intensive algorithms that are primarily limited by processing technology." Because of its design, the M-code signal enables less complex direct acquisition than the P(Y) code. Also, advances in semiconductor technology allow the required processing circuitry and storage to be smaller, less expensive, and to use less power than was possible when limited direct acquisition capability was developed for P(Y) code in the 1990s. But could anyone bring these pieces together in time so that the M-code signal design could rely on direct acquisition? Enter the "DirAc"—short for "direct acquisition"—chip.
A Gathering of Skills While direct acquisition was a great idea on the surface, there were many questions early on about whether it was practical. Viewed from the perspective of the late 1990s, was it too large a step to rely on direct acquisition to satisfy the well-defined requirements that M-code signal acquisition must satisfy? Would such a chip require too much power for battery-powered receivers carried in the field? Would the heat generated by the circuit be excessive? Would each chip cost hundreds or even thousands of dollars? Even if the required chip-manufacturing capability were to exist in time, would it be a non-U.S. facility, which would introduce security concerns? MITRE's initial look at these challenges indicated that the technology would be ready to meet the system needs. But more than a study was needed to provide the DOD with the necessary level of confidence. "The safe solution would have been to rely on a separate acquisition signal," says John Betz, MITRE Fellow. "But Air Force systems engineers recognized the long-term advantages of direct acquisition and had the vision and mettle to invest in building an integrated circuit." MITRE was already developing a test receiver for the modernized signals, and this test receiver needed an acquisition circuit, so the development would satisfy multiple objectives. The MITRE DirAc team included system engineers, mathematicians, signal processors, and electronics engineers. Fite worked on the high-level architecture and signal processing algorithms—such as what computations needed to occur in what order—while Paul Capozza, senior principal integrated electronics engineer, defined the detailed chip architecture and led a team of chip designers in the final design and build of the DirAc chip. Betz performed theoretical performance assessments and systems engineering, while overseeing the integrated effort. "It was the perfect marriage of different skill sets," Fite says. Despite the seemingly insurmountable technical obstacles, the project progressed swiftly. Betz adds, "We all had a 'can do' attitude—including the Air Force personnel who funded and monitored the project." The team also had fun along the way, developing a logo for the chip that portrays a fundamental function in signal processing, the Dirac delta function (named for the British physicist Paul Dirac), overlaying the distinctive spectrum of the M-code signal. Fast Track to a Fast Chip One thing that made the project flow so smoothly was MITRE's in-house capacity for creating chips. "We have a tremendous capability to design integrated circuits for signal processing, GPS, and communications," Capozza explains. "We have a full set of computer-aided design tools, as well as access to foundries and process technology. We've been building chips since the early 1980s and have done a number of designs, including two for the GPS Joint Program Office. So we can merge our implementation and analysis capabilities to provide insights to the program. "Thanks to all these capabilities, it only took about nine months to design and build the chip. We submitted it for fabrication to a foundry and got the prototypes back in about three months. That's a pretty short timeframe from initial concept to a working sample," Capozza adds. The chip's development wasn't the only thing that moved fast. "The prototype worked 10 times faster than anyone ever thought of achieving—which was 100 times faster than any similar-purposed chip at the time," Fite says. "John Betz asked us to make a leap forward, so that's what we were aiming for. We took some risks, but also did a lot of cross-checking along the way." That crosschecking involved much more than a cursory examination of the prototype DirAc chip. "We tested it using 150 million test vectors," Capozza says. "We demonstrated that it worked properly in all operational modes and in different environmental conditions. In a year's timeframe, we produced a complete working device that demonstrated all the various requirements the Joint Program Office wanted to understand." MITRE isn't in the business of manufacturing computer chips, of course. Fite emphasizes that our role was to deliver a working, reproducible prototype to the GPS Joint Program Office. "We wanted to show it could be done, to fully document what we had done, then to let the competing vendors choose how to build upon it. Our goal was not to show the only possible solution, but to set the bar." At each step of the process—from design and testing through the trial manufacturing runs—the team released technical documentation to the contractor community competing to build M-code receivers. Since this information, as well as licenses to use the MITRE-patented DirAc technology, is available to the contractors at no cost, it helps get everyone up to speed faster. The hope is that the government can get better technology, cost effectively. "We know that there are lots of smart designers out there, and that semiconductor technology has already advanced from what we used," says Betz, "so we expect production M-code receivers to include even more capable circuits for direct acquisition." In the meantime, the team continues testing the chip by integrating it into on an M-code receiver testbed at a MITRE laboratory. Keeping It Real Because the DOD isn't requiring its vendors to produce actual production receivers until 2008, DirAc came along early enough to have significant impact on what the system design will look like, both at the space and user/equipment levels. This gives the DirAc team a lot of satisfaction. "It's been exciting to be able to do something that could have a lot of influence on the system," Fite says. "And most important, we showed it really could be done by building it. It's not just on paper—it's real." —by Alison Stern-Dunyak Tackling the direct acquisition challenge took a team of MITRE employees, all working together to conquer several seemingly intractable technical tasks. The team consisted of: John Betz, Thomas Bielicki, Jeffrey Blanchard, Paul Capozza, Brian Faull, John Fite, Michael Fitzgibbons, Samuel Girgis, Brian Holland, Roberto Landrau, and Nimit Nguansiri. Related Information Articles and News
Technical Papers and Presentations |
||||||||||||||||||||
| Page last updated: December 3, 2004 | Top of page |
||||||||||||||||||||
Solutions That Make a Difference.® |
|
|
